IEEE Sensors Journal最新文献

{"title":"TinyFL_HKD: Enhancing Edge AI Federated Learning With Hierarchical Knowledge Distillation Framework","authors":"Chung-Wen Hung;Cheng-Yu Tsai;Chun-Chieh Wang;Ching-Hung Lee","doi":"10.1109/JSEN.2025.3544861","DOIUrl":"https://doi.org/10.1109/JSEN.2025.3544861","url":null,"abstract":"With the rapid evolution of artificial intelligence (AI) and the Internet of Things (IoT), machine learning is increasingly being integrated into embedded systems, bringing computational capabilities closer to where data are generated. This article introduces a tiny federated learning framework, which concerns privacy, personalized training, and the constrained computational resources of edge platforms by introducing a novel hierarchical knowledge distillation (HKD), called TinyFL_HKD. The HKD leverages hierarchical learning and advanced encryption security (AES) schemes to ensure data privacy and security. It employs knowledge distillation to reduce model complexity for implementation in edge devices and enhance personalization. The performance of TinyFL_HKD is introduced by using two datasets: the tool wear dataset and the PHM 2010 Data Challenge dataset. Experimental results indicate that the HKD framework surpasses traditional federated averaging (FedAvg) and personalized federated learning (PFL) algorithms in both model accuracy and computational efficiency. This establishes HKD as a resilient solution for edge AI applications.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"25 7","pages":"12038-12047"},"PeriodicalIF":4.3,"publicationDate":"2025-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143748812","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Ultrafast Humidity Interrogation Based on Agarose-Coated Fiber Bragg Grating With Time-Stretching Method","authors":"Haichao Han;Qiang Ling;Si Luo;Jinghong Zhang;Yihao He;Rujun Zhou;Yunlian Ding;Ding Mao;Anping Xiao;Yusheng Zhang;Zhangwei Yu;Zuguang Guan;Daru Chen","doi":"10.1109/JSEN.2025.3545271","DOIUrl":"https://doi.org/10.1109/JSEN.2025.3545271","url":null,"abstract":"Ultrafast humidity interrogation is valuable in industrial production, healthcare, and environmental monitoring due to its fast response and high sensitivity. We present an ultrafast humidity sensor based on an agarose-coated fiber Bragg grating (AG-FBG) structure combined with a time-stretching method. Leveraging AG’s ability to swell or contract in response to humidity variations results in changes to the effective grating period, which in turn causes a shift in the Bragg wavelength. Using a homemade high-power femtosecond all-polarization-maintaining fiber laser, the time-stretching technique can be achieved through a long dispersion-compensating fiber. According to wavelength-to-time mapping, the wavelength shift associated with humidity response can be converted into the time domain, resulting in a humidity sensitivity of −0.017 ns/% relative humidity (RH) at an interrogation speed of 76.75 MHz and a humidity resolution of 0.007%RH. The proposed humidity monitoring system has demonstrated high stability and repeatability. Therefore, with the advantages of high sensitivity, high accuracy, and ultrafast response time, our proposed humidity sensor shows significant potential in the field of real-time monitoring.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"25 7","pages":"11148-11154"},"PeriodicalIF":4.3,"publicationDate":"2025-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143748965","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Localization of Moving Target in Unknown Complex Background for Single-Pixel Imaging","authors":"Qing-Fan Wu;Peng-Cheng Ji;Hui-Juan Zhang;Shuai-Jun Zhou;Zhao-Hua Yang;Yuan-Jin Yu","doi":"10.1109/JSEN.2025.3544704","DOIUrl":"https://doi.org/10.1109/JSEN.2025.3544704","url":null,"abstract":"Fast target localization in unknown complex backgrounds remains a key challenge for single-pixel imaging (SPI), as existing methods rely heavily on preknown scene information. We propose a novel localization method based on the generalized S-transform (GST) slices of 1-D projections. The GST results correspond to the correlation between the window function at different positions and the scene projection. By selecting appropriate parameters in the initial, the window function can be optimized to closely match the shape and size of the target projection, resulting in a higher response at the target position. The location of the peak response within the sampling area is designated as the target position. This method enables effective localization for different target sizes and unknown complex backgrounds by adjusting the relevant parameters. For a <inline-formula> <tex-math>${256} times {256}$ </tex-math></inline-formula> size scene with a <inline-formula> <tex-math>${48} times {48}$ </tex-math></inline-formula> size target, the simulations and experiments validate that our method improves the localization accuracy and reduces the number of patterns by <inline-formula> <tex-math>$({15}/{16})$ </tex-math></inline-formula> compared to the differential Hadamard projection method using background subtraction. However, the root mean square error of the experiment results was improved by up to 0.3. Furthermore, in order to select appropriate parameters, we also analyzed the influence of different frequencies, object sizes, and sampling region lengths on the localization results.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"25 7","pages":"11782-11791"},"PeriodicalIF":4.3,"publicationDate":"2025-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143761384","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An Ultrahigh-Precision Thin-Film Platinum Resistance Sensor for Ocean Temperature Measurements","authors":"Meng Li;Yonghua Wang;Yuzhen Guo;Yanan Qiao;Shiqiang Zhang;Jianwei Liu;Zengxing Zhang;Danfeng Cui;Yi Chen;Dan Liu;Chenyang Xue","doi":"10.1109/JSEN.2025.3544310","DOIUrl":"https://doi.org/10.1109/JSEN.2025.3544310","url":null,"abstract":"This study presents the design and fabrication of an ultrahigh-precision platinum thin-film temperature sensor without an adhesion layer, which effectively eliminates the performance degradation caused by adhesion layer diffusion during high-temperature annealing. A 1-<inline-formula> <tex-math>$mu $ </tex-math></inline-formula>m-thick platinum film was deposited on an alumina substrate via segmented magnetron sputtering, followed by annealing in air at temperatures ranging from <inline-formula> <tex-math>$500~^{circ }$ </tex-math></inline-formula>C to <inline-formula> <tex-math>$900~^{circ }$ </tex-math></inline-formula>C. The resistance-temperature relationships of the sensor were characterized over the <inline-formula> <tex-math>$0~^{circ }$ </tex-math></inline-formula>C–<inline-formula> <tex-math>$35~^{circ }$ </tex-math></inline-formula>C range under various annealing conditions, with a particular focus on the temperature coefficient of resistance (TCR). The morphological and grain size analyses of the platinum film were conducted using X-ray diffraction (XRD) and scanning electron microscopy. Results indicate that optim al sensor performance was achieved with an annealing temperature of <inline-formula> <tex-math>$800~^{circ }$ </tex-math></inline-formula>C for 2 h, which leads to an increase in the TCR from <inline-formula> <tex-math>$2.36 times 10^{-{3}}$ </tex-math></inline-formula>/°C to <inline-formula> <tex-math>$3.65 times 10^{-{3}}$ </tex-math></inline-formula>/°C. Precision calibration and stability tests show that the sensor achieved an excellent measurement accuracy of <inline-formula> <tex-math>$0.0019~^{circ }$ </tex-math></inline-formula>C and a maximum temperature drift of only <inline-formula> <tex-math>$0.0009~^{circ }$ </tex-math></inline-formula>C per month over a six-month period. These results indicate that the platinum thin-film sensor exhibits outstanding performance, making it particularly suitable for high-precision ocean temperature measurement applications.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"25 7","pages":"10628-10636"},"PeriodicalIF":4.3,"publicationDate":"2025-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143748964","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"SIAR: Signing in the Air Based on Patching-Gated Recurrent Unit Model for Fine-Grained Trajectory Feature Learning","authors":"Shih-Hsiung Lee;Hsuan-Chih Ku","doi":"10.1109/JSEN.2025.3544609","DOIUrl":"https://doi.org/10.1109/JSEN.2025.3544609","url":null,"abstract":"In the current era, where information security and personal privacy are paramount, biometric recognition technology has raised significant privacy concerns. In particular, the COVID-19 pandemic has accelerated the demand for contactless biometric recognition technology, among which facial recognition technology is widely used owing to its convenience, but also faces privacy protection controversies. To address this challenge, this study innovatively proposes a noncontact identity verification method based on hand-air–writing trajectories. This method avoids reliance on personal facial data and reduces the risk of privacy infringement. It also significantly reduces the chance of virus transmission as it requires no contact with any surface, providing a secure and privacy-protective new option for biometric recognition technology. This study adopted the gated recurrent unit (GRU) model to process and analyze the hand trajectory data. The GRU model was selected for its excellent efficiency in handling time-series data and for effectively extracting distinctive features from user hand gestures. Applying this model enabled the system to accurately identify different symbols written in the air by users, thereby completing the identity verification process. To achieve wider application and reduce system operating costs, this study deployed a model on edge-computing devices. At a precision level of fp32, the system achieved an accuracy of 97.33%, demonstrating considerable practical potential.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"25 7","pages":"11772-11781"},"PeriodicalIF":4.3,"publicationDate":"2025-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143761547","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Zero-Shot Defect Detection With Anomaly Attribute Awareness via Textual Domain Bridge","authors":"Zhe Zhang;Shu Chen;Jian Huang;Jie Ma","doi":"10.1109/JSEN.2025.3544407","DOIUrl":"https://doi.org/10.1109/JSEN.2025.3544407","url":null,"abstract":"Visual defect detection is crucial for industrial quality control in intelligent manufacturing. Previous research requires target-specific data to train the model for each inspection task. However, due to the challenges of collecting proprietary data and model-training time costs, zero-shot defect detection (ZSDD) has become an emerging topic in the field. ZSDD, which requires models trained with auxiliary data, can detect defects on different products without target-data training. Recently, large pretrained vision-language models (VLMs), such as contrastive language-image pre-training model (CLIP), have demonstrated revolutionary generality with competitive zero-shot performance across various downstream tasks. However, VLMs have limitations in defect detection, which are designed to focus on identifying category semantics of the objects rather than sensing object attributes (defective/nondefective). The current VLMs-based ZSDD methods require manually crafted text prompts to guide the discovery of anomaly attributes. In this article, we propose a novel ZSDD method, namely attribute-aware CLIP, to adapt CLIP for anomaly attribute discovery without designing specific textual prompts. The core is designing a textual domain bridge, which transforms simple general textual prompt features into prompt embeddings better aligned with the attribute awareness. This enables the model to perceive the attributes of objects by text-image feature matching, bridging the gap between object semantic recognition and attribute discovery. Additionally, we perform component clustering on the images to break down the overall object semantics, encouraging the model to focus on attribute awareness. Extensive experiments on 16 real-world defect datasets demonstrate that our method achieves state-of-the-art (SOTA) ZSDD performance in diverse class-semantic datasets.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"25 7","pages":"11759-11771"},"PeriodicalIF":4.3,"publicationDate":"2025-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143761368","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Skeletal Sensor With 2-D Detection Capability","authors":"Limin Ren;Xu Zhang;Yisong Tan","doi":"10.1109/JSEN.2025.3544316","DOIUrl":"https://doi.org/10.1109/JSEN.2025.3544316","url":null,"abstract":"Currently, existing skeletal sensors suffer from a 1-D detection, complex detection systems, and low detection sensitivity. In this article, a skeletal sensor with 2-D detection is proposed. The sensor consists of a substructure and detection elements. The substructure is composed of a negative Poisson’s ratio structure to increase sensor sensitivity. The detection element is composed of magnetostrictive material 2826 MB. Helmholtz coils and planar coils are used for reducing the detection complexity. A theoretical model of the interaction between the underlying structure and the magnetostrictive material is established. The sensor prototype is processed. Experimental platforms are set up for comprehensive testing of the sensor. Experimental results show that the sensor can perform 2-D force detection; it has a sensitivity up to 0.0381 mV/N and 0.6748 mV/N<inline-formula> <tex-math>$cdot $ </tex-math></inline-formula>m. The maximum hysteresis error is only 2.659%, and its maximum load is 4.849 kN and 48.14 N<inline-formula> <tex-math>$cdot $ </tex-math></inline-formula>m. The maximum load of the sensor fully meets people’s daily life needs. The sensor system does not require complicated operations. All the above factors indicate that the proposed sensor is a good solution to the problems faced by skeletal sensors currently, and it has potential applications in future skeleton research.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"25 7","pages":"12016-12024"},"PeriodicalIF":4.3,"publicationDate":"2025-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143748967","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Noncontact Radar Sensing of Wrist Pulse Wave With Enhanced Accuracy and Flexibility","authors":"Xianzhong Tian;Haotian Shi;Yifan Gao;Xiaoyu Zhang;Yongxin Guo","doi":"10.1109/JSEN.2025.3543919","DOIUrl":"https://doi.org/10.1109/JSEN.2025.3543919","url":null,"abstract":"Wrist pulse wave (WPW) is an important vital sign signal for digital health. By recovering WPW in a noncontact manner with wearable sensors could provide a new approach for the diagnostics and prediction of various human conditions and diseases. To this end, this work proposes a self-injection locking (SIL) principle-based wrist-worn radar sensor with enhanced accuracy and flexibility. By accuracy, this sensor is advantageous for the high correlation coefficients compared to a contact sensor in the time domain. By flexibility, it can accommodate a variety of microwave antennas, enabling a range of related applications. In addition, the instability problem of the traditional SIL radar is analyzed jointly considering the conditions of the oscillator and the pulse signal, concluding a possible solution to the instability problem. Experiments are conducted by comparing the WPW data simultaneously collected by the fabricated 5.8 GHz sensor and a contact piezoelectric sensor. High correlation coefficients are observed at two typical sensor-skin distances: an average of 0.841 at 1 mm and 0.773 at 5 mm, showing high accuracy. This work could facilitate research works for a range of related biomedical applications.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"25 7","pages":"11995-12002"},"PeriodicalIF":4.3,"publicationDate":"2025-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143748876","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Factor Graph Optimization for Robust Indoor Positioning: A Data-Driven Approach Integrating Audio Ranging and Pedestrian Dead Reckoning","authors":"Wangdi Ke;Ruizhi Chen;Lixiong Huang;Guangyi Guo","doi":"10.1109/JSEN.2025.3544586","DOIUrl":"https://doi.org/10.1109/JSEN.2025.3544586","url":null,"abstract":"This work presents a novel approach for indoor positioning by integrating a data-driven audio ranging algorithm with pedestrian dead reckoning (PDR) constrained by magnetic information (MI). The proposed system leverages convolutional neural networks (CNNs) to process time-domain audio signals by transforming them into spectrograms, thus enhancing the accuracy of signal arrival time estimation in complex indoor environments. The PDR system operates at 20 Hz and adapts to various smartphone usage postures by combining sensor data from microphones, BLE, and IMU sensors. In order to improve robustness, the proposed system incorporates multiple robust factors within particle filter (PF) and factor graph optimization (FGO) algorithms, thus effectively mitigating abnormal observations and reducing positioning errors. The experimental results demonstrate that the proposed system achieves high positioning accuracy, with 95% of errors being within 1 m and maximum errors not exceeding 1.7 m across different smartphones, making it a viable solution for precise indoor positioning in real-world scenarios.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"25 7","pages":"12025-12037"},"PeriodicalIF":4.3,"publicationDate":"2025-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143748782","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Novel Solution for Solving Time-Varying Algebraic Riccati Equations and Its Application to Sound Source Tracking","authors":"Chuncheng Chen;Zhiyuan Song;Keer Wu;Kaixiang Yang;Xiuchun Xiao","doi":"10.1109/JSEN.2025.3538173","DOIUrl":"https://doi.org/10.1109/JSEN.2025.3538173","url":null,"abstract":"Solving time-varying algebraic Riccati equations (TVAREs) is crucial in sound source tracking and optimal control. It is worth noting that previous studies have focused primarily on solving static algebraic Riccati equations (AREs) or interference-free TVAREs. Nonetheless, in real-world solution systems, AREs are often time-varying and subject to a variety of external disturbances. To address these problems, we propose two strong initial state discrete noise-resistant zeroing neurodynamics (SDRZND) algorithms for determining the solutions to TVAREs. First, we introduce a bounded smoothing of the strong initial state coefficient to accelerate algorithm convergence while avoiding the additional impulse noise that nonsmoothed coefficients in the discrete algorithm might generate. Then, an integral feedback term is designed and integrated with this coefficient to enhance the algorithm’s robustness. Subsequently, to further improve the algorithm’s flexibility, we introduce a variable time step, leading to the SDRZND-Euler (SDRZND-E) and SDRZND-Taylor–Zhang (SDRZND-TZ) algorithms presented in this article. Lastly, the effectiveness, noise resistance, and practicality of these algorithms are verified through theoretical analysis, numerical simulations, and sound source tracking experiments.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"25 7","pages":"11155-11166"},"PeriodicalIF":4.3,"publicationDate":"2025-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143748811","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}